We Keep Coding

Regular Expression that matches on values after a pipe in between brackets

I'm still learning a lot about regex, so please forgive any naivety.
I've been using this site to test:
http://www.systemtextregularexpressions.com/regex.match
Basically, I'm having issues writing a regular expression that will match on any value after a pipe in between brackets.
Given an example string of:
"<div> \n [dont1.dont2|match1|match2] |dont3 [dont4] dont5. \n </div>"
Expected output would be a collection:
match1,
match2
The closest I've been able to get so far is:
(?!\[.*(\|)\])(?:\|)([\w-_.,:']*)
Above gives me the values, including the pipes, and dont3.
I've also tried this guy:
\|(.*(?=\]))
but it outputs:
|match1|match2

Here's one way of doing it:
(?<=\[[^\]]*\|)[^\]|]*
Here's the meaning of the pattern:
(?<=\[[^\]]*\|) - Lookbehind expression to ensure that any match must be preceded by an open bracket, followed by any number of non-close-bracket characters, followed by a pipe character
(?<= ... ) - Declares a lookbehind expression. Something matching the lookbehind must immediately precede the text in order for it the match. However, the part matched by the lookbehind is not included in the resulting match.
\[ - Matches an open bracket character
[^\]]* - Matches any number of non-close-bracket characters
\| - Matches a pipe character
[^\]|]* - Matches any number of characters which are neither close brackets nor pipe characters.
The lookbehind is greedy, so it will allow for any number of pipes between the open bracket and the matching text.

try this:
\[.*?(?:\|(?<mydata>.*?))+\]
note: the online tool will only show you the last capture inside a quantifed () for a given match, but .NET will remember each capture of a group that matches multiple times

Try this:
^<div>\s*[^|]+|([^|]+)|([^|]+)

RegEx match on any of multiple groups

I'm not sure if this is possible, but I would like to match on multiple regex groups
(^[0-9]) (^[$][0-9]) (^[$]{2}[0-9])
It would match the string if the first character is number, or if the first character is a $ followed by a number, or if the first two characters are a $ followed by a number.
Example strings that would match:
15271%
$3C001%
$$8244150928223C001%
Can this be done in one go, or would I have to check each match individually?
Any help is appreciated. Thanks!

You can make make use of the pipe symbole | to achieve that. It basically behaves like an "or" in your regex pattern.
For example:
(banana|apple)
would match both "banana" and "apple".
In your case, you can also use a pattern like this
(\${0,2}\d.+)
to match all options: without $, with one $ and with two $.

You could use:
^\d.*|^\$\d.*|^\$\$\d.*
try {
if (Regex.IsMatch(subjectString, #"\A(?:^\d.*|^\$\d.*|^\$\$\d.*)\z", RegexOptions.Multiline)) {
// Successful match
} else {
// Match attempt failed
}
} catch (ArgumentException ex) {
// Syntax error in the regular expression
}

regular expression ".*[^a-zA-Z0-9_].*"

As I am trying to read more about regular expressions in C#, I just want to make sure of my conclusion that I made.
for the following expression ".*[^a-zA-Z0-9_].* ", the " .* " at the beginning and end are useless, is that right ? because as I understood, that ".*" means zero or more occurrence of any character, but being followed by "[^a-zA-Z0-9_]" which means any character other than any combination of letters and digits case insensitive, makes ".*" useless to be added before and after "[^a-zA-Z0-9_]", is that right ?
Here is the code I am using to check if the expressions matches
// Here we call Regex.Match.
Match match = Regex.Match("anytest#", ".*[^a-z A-Z0-9_].*");
//Match match = Regex.Match("anytest#", "[^a-z A-Z0-9_]");
// Here we check the Match instance.
if (match.Success)
Console.WriteLine("error");
else
Console.WriteLine("no error");

.*[^a-zA-Z0-9_].* will match the entire input as long as there is a non-alphanumeric/underscore somewhere in the input. [^a-zA-Z0-9_] will match only a single non-alphanumeric/underscore character (most likely the last one, if you're using the default greedy matching) if it is somewhere in the input. Which one you want depends on the input and what you want to do once you find out if a non-alphanumeric/underscore character exists in the input.

The only difference would be whether the "margin characters" will be included in the result or not.
For:
ab41--_71j
It will match:
1--_7
And without the .* at beginning and end it will match:
--_
Any string will match the .*[^a-zA-Z0-9_].* regex at least once as long as it has at least one character that isn't a-zA-Z0-9_
From your currently last comment in your answer, I understand that you actually use:
^[a-zA-Z0-9]*$
This will match only if all characters are digit/letters.
If it doesn't match, then the string is invalid.
If you also want to allow the _ character, then use:
^[a-zA-Z0-9_]*$
Which can even be shortened to:
^\w$
In general, it is better to make regex's Validate rather than Invalidate strings. It just makes more sense and is more intuitive.
So my validation would look like:
if (Regex.IsMatch("anytest#", "^\\w$"))
{
Console.WriteLine("Success");
}
else
{
Console.WriteLine("Error");
}
Another option that is probably faster:
if ("anytest#".ToCharArray().All(c => char.IsLetterOrDigit(c) || c == '_'))
{
Console.WriteLine("Success");
}
else
{
Console.WriteLine("Error");
}
And if you don't want '_' to be included, it can even look nicer;
if ("anytest#".ToCharArray().All(char.IsLetterOrDigit))
{
Console.WriteLine("Success");
}
else
{
Console.WriteLine("Error");
}

No, because there are other characters than a-Z and 0-9.
That regex matches all strings that start with any characters followed not by a-zA-Z0-9 and end with any characters. Or just a string that does not contain a-zA-Z0-9 at all.
If you leave the .* then you just have a regex that matches a charatcer that does not contain a-zA-Z0-9 at all.
.*[^a-zA-Z0-9_].* matches for instance: ABC_ß_ABC
[^a-zA-Z0-9_] matches for instance: ß (and this regex just matches 1 character)

Input 1 : ABC_ß_ABC
Input 2 : ß
Regex 1: .*[^a-zA-Z0-9_].*
Regex 2: [^a-zA-Z0-9_]
Both the inputs match both the regex,
For input 1
Regex 1 matches 9 characters
Regex 2 matches only 1 character

Only include those tokens in the Regex that you are actually looking for. In your case you didn't actually care whether there are any other characters before or after the excluding character class you specified. Adding .* before and after that doesn't change the success of the match, but makes matching more complicated. A Regex matches anywhere already, unless you specifically anchor it somehow, e.g. using ^ at the start.

Problem with regex, how do I get all with \S up until a special character?

Ive got the text:
192.168.20.31 Url=/flash/56553550_hi.mp4?token=(uniquePlayerReference=81781956||videoId=1)
And im trying to get the uniquePlayerReference and the videoId
Ive tried this regular expression:
(?<=uniquePlayerReference=)\S*
but it matches:
81781956||videoId=1)
And then I try and get the video id with this:
(?<=videoId=)\S*
But it matches the ) after the videoId.
My question is two fold:
1) How do I use the \S character and get it to stop at a character? (essentially what is the regex to do what i want) I cant get it to stop at a defined character, I think I need to use a positive lookahead to match but not include the double pipe).
2) When should I use brackets?

The problem is the mul;tiplicity operator you have here - the * - which means "as many as possible". If you have an explicit number in mind you can use the operator {a,b} where a is a minimum and b a maximum number fo matches, but if you have an unknown number, you can't use \S (which is too generic).
As for brackets, if you mean () you use them to capture a part of a match for backreferencing. Bit complicated, think you need to use a reference for that.
I think you want something like this:
/uniquePlayerReference=(\d+)||videoId=(\d+)/i
and then backreference to \1 and \2 respectively.

Given that both id's are numeric you are probably better off using \d instead of \S. \d only matches numeric digits whereas \S matches any non-whitespace character.
What you might also do is a non gready match up till the character you do not want to match like so:
uniquePlayerReference=(.*?)\|\|videoId=(.*?)\)
Note that I have escaped both the | and ) characters because otherwise they would have a special meaning inside a regex.
In C# you would use this like so: (which also answers your question what the brackets are for, they are meant to capture parts of the matched result).
Regex regex = new Regex(#"uniquePlayerReference=(.*?)\|\|videoId=(.*?)\)");
Match match = regex.Match(
"192.168.20.31 Url=/flash/56553550_hi.mp4?token=(uniquePlayerReference=81781956||videoId=1)");
if (match.Success)
{
string playerReference = match.Groups[1].Value;
string videoId = match.Groups[2].Value;
// Etc.
}

If the ID isn't just digits then you could use [^|] instead of \S, i.e.
(?<=uniquePlayerReference=)[^|]*
Then you can use
(?<=videoId=)[^)]*
For the video ID

The \S means it matches any non-whitespace character, including the closing parenthesis. So if you had to use \S, you would have to explicitly say stop at the closing parenthesis, like this:
videoId=(\S+)\)
Therefore, you are better off using the \d, since what you are looking for are numeric:
uniquePlayerReference=(\d+)
videoId=(\d+)

C# regex to replace a delimiter by another one

I'm working on pl/sql code where i want to replace ';' which is commented with '~'.
e.g.
If i have a code as:
--comment 1 with;
select id from t_id;
--comment 2 with ;
select name from t_id;
/*comment 3
with ;*/
Then i want my result text as:
--comment 1 with~
select id from t_id;
--comment 2 with ~
select name from t_id;
/*comment 3
with ~*/
Can it be done using regex in C#?

Regular expression:
((?:--|/\*)[^~]*)~(\*/)?
C# code to use it:
string code = "all that text of yours";
Regex regex = new Regex(#"((?:--|/\*)[^~]*)~(\*/)?", RegexOptions.Multiline);
result = regex.Replace(code, "$1;$2");
Not tested with C#, but the regular expression and the replacement works in RegexBuddy with your text =)
Note: I am not a very brilliant regular expression writer, so it could probably have been written better. But it works. And handles both your cases with one-liner-comments starting with -- and also the multiline ones with /* */
Edit: Read your comment to the other answer, so removed the ^ anchor, so that it takes care of comments not starting on a new line as well.
Edit 2: Figured it could be simplified a bit. Also found it works fine without the ending $ anchor as well.
Explanation:
// ((?:--|/\*)[^~]*)~(\*/)?
//
// Options: ^ and $ match at line breaks
//
// Match the regular expression below and capture its match into backreference number 1 «((?:--|/\*)[^~]*)»
// Match the regular expression below «(?:--|/\*)»
// Match either the regular expression below (attempting the next alternative only if this one fails) «--»
// Match the characters “--” literally «--»
// Or match regular expression number 2 below (the entire group fails if this one fails to match) «/\*»
// Match the character “/” literally «/»
// Match the character “*” literally «\*»
// Match any character that is NOT a “~” «[^~]*»
// Between zero and unlimited times, as many times as possible, giving back as needed (greedy) «*»
// Match the character “~” literally «~»
// Match the regular expression below and capture its match into backreference number 2 «(\*/)?»
// Between zero and one times, as many times as possible, giving back as needed (greedy) «?»
// Match the character “*” literally «\*»
// Match the character “/” literally «/»

A regex is not really needed - you can iterate on lines, locate the lines starting with "--" and replace ";" with "~" on them.
String.StartsWith("--") - Determines whether the beginning of an instance of String matches a specified string.
String.Replace(";", "~") - Returns a new string in which all occurrences of a specified Unicode character or String in this instance are replaced with another specified Unicode character or String.